Most senior leaders would agree that adapting in real time to customer and market needs is vital for achieving their visions and goals. And to develop this capability, they’d also likely agree they need data and artificial intelligence (AI).
The examples are all around us. The city of Chicago uses 12 variables, including high daily temperatures, to prioritize which of the city’s 7,000 “high-risk” restaurants it should send its 35 food inspectors to. The AI solution found violations a week earlier than they otherwise would have.
We helped an insurer create an AI system that provides underwriters with more precise estimates of risk, as well as the likelihood applicants will accept a specific price quote. The insurer can adjust how the AI environment makes its approval and pricing recommendations to ensure compliance with changing corporate priorities around risk vs. revenue.
We’ve also seen AI help predict customers’ emerging needs as markets change. If the economy goes into a recession, informed analysis could help organizations not only cut costs but also provide the products and services customers might need in a downturn.
And, of course, there are the businesses that have used data and AI to create new revenue streams and business models that drive lasting competitive advantage. For game changers like Uber, data is at the core of the company and constitutes value, not cost.
Facing Up to AI Realities
But there’s another hard fact that would draw consensus from many business leaders today: the unprecedented effort required to move an enterprise in an AI direction. The fact is, AI-enabled business change requires as much alteration to corporate culture, organizational structures and processes, and workforce roles and skills as it does new technology.
For example, too many organizations still treat data as an expense and a security risk. Technical or organizational siloes make it difficult to pool information in flexible data lakes. Many businesses also lack the skills to manage AI-enabled analytics amid rising privacy and ethical concerns. Others are unable to provide audit trails on how AI decisions were made or deliver AI-enabled analytics quickly enough.
AI Success Factors
Here are a few common characteristics we’ve seen among businesses that have realized success with data and AI:
- Senior leadership is passionate about the value of data. CIOs can play a big role here. They need to work with the CEO and business leaders to resolve the tension between business managers who want more access to data and the pressure to restrict such access to reduce cost and risk.
- Data scientists embedded with business users. By working more closely with business users, data scientists can better understand the business context behind their requests. Business managers may say they need “X,” but what they really need is a solution to a problem, which may or may not be “X.”
- The establishment of an AI oversight role. Because AI systems learn and improve, they can produce different results at different times based on the data they’re given and the algorithms applied. This means they require closer oversight than traditional systems that are coded, tested and summarily released – and then left alone.
In some ways, managing an AI system is like raising a child. You need to be sure they’re not being trained with intentionally or unintentionally biased data or by malicious users. If a loan application is denied, for example, you need to know the decision was ethically sound.
- Plenty of high-quality data. The more high-quality data, the better and more quickly the AI system can learn. To supply that data, organizations must refine their processes for ongoing data preparation, integration and pipelining. This essential groundwork is the hardest part in building an AI system.
- Modernized infrastructure. An infrastructure based on application programming interfaces (APIs) and Agile development techniques makes it easier to quickly deliver new AI-based applications. This more flexible infrastructure enables organizations to better access needed data from outside the enterprise, and more effectively monetize the resulting insights by sharing them across the ecosystem.
For instance, we worked with one of the world’s largest and busiest airports to revamp its technology infrastructure to increase airport efficiency and performance. The organization combined data and AI to unlock more capacity to serve airlines and reduce passenger misconnects.
Our final recommendation: Start now. AI can be difficult and complex, and it’s hard to catch up once you’ve fallen behind. Even modest successes will teach you a lot, and the real danger in digital is being a laggard.
This blog was distilled from content delivered during our recent webinar How Organizations Transform Using Data and AI, produced with JP Gowdner, a VP and Principal Analyst with Forrester Research.
Subscribe to our newsletter and get expert insights straight to your inbox.×
SUBSCRIBE TO OUR NEWSLETTER✖
Thanks for your interest in Digitally Cognizant.
To complete the subscription process, please click the link in the email we just sent you.